Oracle AI Release Tracker
The Oracle AI Release Tracker is a chronological timeline of how Oracle Fusion's AI capabilities — AI agents, AI Agent Studio, agentic applications, and the security and interoperability plumbing around them — have moved quarter by quarter, from 24D through 26A and into future direction. It answers a single question: what changed, and when.
This is a what/when artifact, not an impact analysis. Where a quarter's AI content is confirmed on this site, it is named specifically. Where it is not, the row says so plainly rather than inventing feature lists — because a release tracker is only useful if you can trust every cell in it.
What This Tracker Is — and What It Is Not
The tracker is deliberately narrow. It records the timeline of Oracle's AI features across Fusion releases so an ERP, HCM, or QA lead can see, in one table, which quarter introduced which class of AI capability and what that implies for testing. It does not rank changes by business impact, walk through a change-analysis process, or cover the hundreds of non-AI features that ship in every Oracle update.
Scope note — how this differs from two neighbouring pages.
- Vs. Oracle AI Release Intelligence (the how). Oracle AI Release Intelligence is the analysis and impact method — how to read a quarterly update, judge which AI changes matter to your configuration, and turn that into a test plan. This tracker is the what/when: the raw chronology those methods are applied to. Use the tracker to spot the change; use Release Intelligence to decide what to do about it.
- Vs. the general Oracle Release Intelligence hub (all modules). Oracle Release Intelligence tracks every Oracle Fusion module — Payables, Receivables, HCM, SCM, and the rest — across quarterly updates. This page is the AI-only slice of that story: agents, Studio, and agentic apps, nothing else. If you need the full module-by-module picture, start at that hub instead.
If you are new to the underlying concepts, the Oracle AI hub orients you across the whole vertical, Oracle AI Agents explains what agents actually are, and Oracle AI Agent Studio covers the build environment. This page assumes you already know the vocabulary and just want the timeline.
Oracle AI Release Timeline: 24D → Future
Each row is one Oracle Fusion release. Only 26A is itemized with specific, named capabilities, because those are documented on this site's own quarterly pages. Every other quarter is marked "Direction / not itemized here" — meaning the release did contain AI movement, but we do not restate specific feature or agent names for it here. Always confirm any quarter against Oracle's official release notes before you plan against it.
| Release | AI agents | Agent Studio | Marketplace / Agentic apps | Security / other | Testing recommendation |
|---|---|---|---|---|---|
| 24D | Direction / not itemized here | Direction / not itemized here | Direction / not itemized here | See release notes | Baseline embedded-AI behaviour before agentic features; regression-test any AI-assisted flow you rely on. |
| 25A | Direction / not itemized here | Direction / not itemized here | Direction / not itemized here | See release notes | Confirm against Oracle release notes; do not assume specific agents. Re-baseline AI-touched screens. |
| 25B | Direction / not itemized here | Direction / not itemized here | Direction / not itemized here | See release notes | Confirm against Oracle release notes; treat any agent behaviour as version-specific and re-verify. |
| 25C | Direction / not itemized here | Direction / not itemized here | Direction / not itemized here | See release notes | Confirm against Oracle release notes; validate Redwood-page changes affecting AI surfaces. |
| 25D | Direction / not itemized here | Direction / not itemized here | Direction / not itemized here | See release notes | Confirm against Oracle release notes; re-run your AI regression pack before the 26A jump. |
| 26A | Ledger Agent (Agentic AI-powered General Ledger experience); Applicant Screening AI Agent in HCM. See 26A ERP agents and 26A HCM. | AI Agent Studio Redwood integration; REST API access for AI agents. | Agentic Applications Builder — compose AI workflows and reusable agents in natural language. See agentic applications. | MCP and A2A interoperability support for cross-system agent collaboration. | Test new agents for prompt, workflow, API and human-approval behaviour; regression-test GL and recruiting flows the agents touch. |
| 26B | Direction / not itemized here | Direction / not itemized here | Direction / not itemized here | See release notes | Not yet substantiated here; confirm from Oracle release notes when published, then re-plan agent tests. |
| 26C | Direction / not itemized here | Direction / not itemized here | Direction / not itemized here | See release notes | Not yet substantiated here; confirm from Oracle release notes when published. |
| Future | Direction / not itemized here | Direction / not itemized here | Direction / not itemized here | See release notes | Observed trend only, not a confirmed Oracle commitment; keep an evergreen AI regression suite ready. |
Only the 26A row is itemized with named capabilities, all of which are documented on this site's quarterly-update pages. Rows marked "Direction / not itemized here" indicate the tracker does not restate specific features for that quarter — verify against Oracle's official release notes.
How to Read the Tracker
Read the table left to right for a single quarter to see the shape of that release, or top to bottom down one column to watch a capability category evolve. A few conventions make the tracker trustworthy:
Itemized rows
Only 26A names specific capabilities. Those names — Ledger Agent, Applicant Screening AI Agent, Agentic Applications Builder, AI Agent Studio Redwood integration, REST API access, MCP/A2A — come from this site's own 26A pages, so they are safe to plan against.
"Direction / not itemized here"
This label means the quarter carried AI movement, but the tracker deliberately does not restate specific agent or feature names for it. It is an honest gap, not a claim that nothing happened. Confirm specifics in Oracle's release notes.
The five columns
AI agents (business-process agents), Agent Studio (the build and integration layer), Marketplace / Agentic apps (reusable and composable AI), Security / other (access, interoperability, controls), and a per-quarter testing recommendation.
Future rows are directional
Anything past the current release is an observed trend, not a confirmed Oracle commitment. Treat 26B, 26C, and Future as prompts to watch, not fixed scope you can build a test plan on today.
When you need to move from "what changed" to "what to test," the tracker hands off to two places: Oracle AI Testing for the validation discipline overall, and Oracle AI Agent testing for the test types specific to agents. For a structured, searchable view of agents by module and process rather than by quarter, see the Oracle AI Agent Directory.
Why Tracking AI Releases Matters
Traditional Oracle features change on a schedule you can reason about: a page moves, a field is added, a program gains a parameter. AI agents are different. An agent's behaviour is shaped by prompts, models, guardrails, and the data it reads — any of which Oracle can adjust across a quarterly update without changing a screen you would notice. The same agent can screen a candidate, summarise a ledger, or route an approval slightly differently in 26A than it did the quarter before. That is regression risk you cannot see unless you were already watching the timeline.
Three properties of Oracle AI make a quarter-by-quarter record worth keeping:
| Property | Why it creates release risk | What the tracker gives you |
|---|---|---|
| Behaviour can drift silently | Prompt, model, or guardrail changes alter agent output with no visible UI change | A dated record of when agent capabilities entered, so you know which quarter to re-baseline |
| New surfaces expand exposure | REST API access and MCP/A2A let agents act across systems, widening what a regression can touch | A column dedicated to security and interoperability so new access paths are not missed |
| Composition compounds change | Agentic apps built from reusable agents inherit every upstream behaviour change | A marketplace / agentic-apps column that flags when composable building blocks arrive |
Without a tracker, teams discover AI changes reactively — a screening result looks off, a ledger summary reads wrong, an approval routes to the wrong person — and then reverse-engineer which quarter caused it. With a tracker, the timeline is the starting point, and release intelligence turns each entry into a decision about what to re-test.
Recommended Action Each Quarter
Whatever a given quarter contains, the tracker supports a repeatable four-step response. The point is to make AI change a planned event rather than a surprise found in production.
- 1. Confirm the quarter's AI content. Start from the tracker row, then verify specifics against Oracle's official release notes. Do not act on a "Direction / not itemized here" row until you have named the actual changes.
- 2. Assess impact with a method, not a guess. Run each confirmed change through Oracle AI Release Intelligence to judge which agents, Studio changes, or interoperability additions actually touch your configuration and processes.
- 3. Plan the tests that fit the change. Map each in-scope change to test types on Oracle AI Agent testing — prompt, workflow, API, regression, security, and human-approval — within the broader Oracle AI Testing discipline.
- 4. Re-run and record. Execute the plan, capture evidence, and note the result against the quarter so next cycle starts from a known baseline. For a wider, module-by-module view beyond AI, cross-reference the Oracle Release Intelligence hub.
Applied to 26A. The confirmed changes — Ledger Agent, Applicant Screening AI Agent, Agentic Applications Builder, AI Agent Studio Redwood integration, REST API access, and MCP/A2A interoperability — mean this quarter's action list centres on GL and recruiting agent behaviour, the new API and interoperability surfaces, and any agents composed in the Agentic Applications Builder. See the 26A ERP & AI agents and 26A HCM pages for the detail.
How SyntraFlow Helps
SyntraFlow helps organizations connect the timeline in this tracker to concrete test planning. Rather than treating each Oracle quarterly update as an unbounded surprise, teams can use the tracker to anchor a recurring review and then decide, with evidence, which AI changes warrant re-testing.
- →SyntraFlow can be configured to align a regression suite with the AI surfaces a given quarter touches — the agent-driven flows, the Redwood pages, and the API paths a release introduces.
- →The SyntraFlow approach can connect release intelligence with test planning, so a change spotted on the timeline becomes a scoped, prioritised set of agent tests rather than a manual scramble.
- →SyntraFlow helps organizations assess whether an agent still behaves as expected across a quarter boundary — capturing evidence per run so a passing result reflects verified behaviour, not an assumption.
- →Because Oracle's AI surfaces evolve, SyntraFlow's roadmap can support new agent, Studio, and interoperability test patterns as they are confirmed, rather than committing to fixed coverage of features that have not shipped.
To be clear about the boundary: the AI in this tracker is Oracle's AI — the agents and Studio you must govern and test. SyntraFlow's own AI-driven testing engine is a separate capability used to help test Oracle. This page describes how the two connect around the release timeline; it does not claim SyntraFlow automatically covers every Oracle AI agent or guarantees any specific outcome.
Frequently Asked Questions
What is the Oracle AI Release Tracker?
▼
It is a chronological timeline of how Oracle Fusion's AI capabilities — AI agents, AI Agent Studio, agentic applications, and related security and interoperability features — have changed from release 24D through 26A and into future direction. It records what changed and when, so teams can see at a glance which quarter introduced which class of AI capability.
How is this different from Oracle AI Release Intelligence?
▼
The tracker is the what/when; Oracle AI Release Intelligence is the how. This page gives you the raw chronology of AI changes across quarters. Release Intelligence is the analysis and impact method you apply to those changes — judging which ones matter to your configuration and turning them into a test plan. Use them together: spot the change here, decide what to do about it there.
How does it relate to the general Oracle Release Intelligence hub?
▼
The Oracle Release Intelligence hub covers every Oracle Fusion module across quarterly updates. This tracker is the AI-only slice of that story — agents, Studio, and agentic apps, and nothing else. If you need the full module-by-module release picture, start at that hub; if you only care about AI, stay here.
Why is only 26A itemized with specific features?
▼
Because 26A is the quarter whose AI capabilities are documented on this site's own quarterly-update pages, so its named features — Ledger Agent, Applicant Screening AI Agent, Agentic Applications Builder, AI Agent Studio Redwood integration, REST API access, and MCP/A2A interoperability — can be stated with confidence. For other quarters, the tracker uses "Direction / not itemized here" rather than inventing feature or agent names it cannot substantiate.
What does "Direction / not itemized here" mean?
▼
It means the release carried AI movement, but this tracker deliberately does not restate specific agent or feature names for that quarter. It is an honest gap, not a claim that nothing shipped. Confirm the specifics for any such quarter against Oracle's official release notes before you plan tests around it.
Why do AI releases need tracking more than ordinary features?
▼
An agent's behaviour depends on prompts, models, guardrails, and the data it reads — Oracle can adjust any of these across a quarterly update without a visible screen change. That makes AI regression risk harder to see than a moved field or new button. A dated timeline tells you which quarter to re-baseline so behaviour drift is caught deliberately rather than in production.
What should I do when a new quarter introduces AI changes?
▼
Confirm the quarter's AI content against Oracle release notes, assess impact through release intelligence, map in-scope changes to test types on Oracle AI Agent testing within the wider Oracle AI Testing discipline, then re-run and record the result against the quarter as your new baseline.
Does SyntraFlow automatically test every Oracle AI agent in a release?
▼
No — and this page does not claim that. SyntraFlow can be configured to align a regression suite with the AI surfaces a quarter touches, and helps organizations assess whether agents still behave as expected across a release boundary. It connects the timeline to test planning; it does not guarantee coverage of features that have not shipped or outcomes for agents it has not been configured to test.
Review Oracle AI Quarterly Changes
Turn the AI release timeline into a plan. See how SyntraFlow can be configured to align your agent and regression testing with each Oracle quarterly update, from 26A onward.